What do we learn from clitics? The case study of Polish

What do we learn from cliticsThe case study of Polish

Albert Ventayol Boada

Wiener Linguistische GazetteInstitut fuumlr Sprachwissenschaft

Universitaumlt WienAusgabe (2014) 128-147

Abstract

Clitics pose a challenge to any linguistic theory that does not abandon the classical

stratification of grammar in modules Halfway between phonology and syntax

clitics have their own morphological features different from both morphemes and

syntactically independent elements mdash words Revising the described typologies

and taking Polish as a case study our aim is to see whether an on-going process

of grammaticalization of these elements in this particular language exists or not

Bearing this in mind we designed a value judgement test that we spread online

between target groups from 16 to 64 years old The obtained data confirmed our

hypothesis as the age of the participants decreased so did the acceptance of the

mobility of clitics as well as their adjunction to non-verbal hosts A wide range

of consequences derive from these results ranging from the theoretical treatment

of these elements to their computational manipulation and the elaboration of

teaching materials of Polish as a L2

Key words clitic grammaticalization host personal markers Polish

1 Introduction the Polish verbal clitics

In recent years Polish has been the object of research in the area of clitics not only because

of the large quantity of clitics found in the language but also because of their variety and their

atypical behaviour within the Slavic family At this point we consider a relevant guide the

synthesis that Franks amp King (2000) developed on the theory of Polish clitics under the gaze of

our practical and theoretical knowledge of the language

The first distinction that these authors make regarding Polish clitics is based on their

functionality in this sense they distinguish between verbal nominal and other clitics which

128

What do we learn from clitics The case study of Polish 129

Preterite (clitic form) Copula (full form)

1st p sg -(e)m jestem2nd p sg -(e)s jestes3rd p sg mdash jest1st p pl -(e)smy jestesmy2nd p pl -(e)scie jestescie3rd p pl mdash sa

Table 1 Verbal clitics in Polish

we consider discursive clitics following the terminology of Spencer amp Luiacutes (2012) However

here we will only revise the first group Table 1 adapted from the same authors summarises the

existing verbal clitics

There are several points in the above table that must be kept in mind First of all the

dichotomy between preterite and copula lies in two facts On the one hand the clitic forms

appear only in correlation with the past tenses of verbs on the other hand the similarity between

verbal clitics and the conjugated forms of the verb byc lsquoto bersquo cannot be mere coincidence In a

general sense it should be kept in mind that the clitics shown in the table act as person markers

ie they introduce information about number and person except for the third persons which

do not have specific forms Finally regarding the e in parentheses it is normally identified as a

marker of the masculine gender1 This however is inaccurate because the nature of this vowel

is epenthetic (Swan 2002)

(1) (a) Ja czytał-e=m te ksiazke

[I readPRF -M=1SG thisACC bookACC]

lsquoI read this bookrsquo

(b) Ja=m czytał-Oslash te ksiazke

[I=1SG readPRF-M thisACC bookACC]


(c) On czytał-Oslash te ksiazke

[He readPRF-M thisACC bookACC]

(d) Ona czytał-a te ksiazke

[She readPRF-F thisACC bookACC]

lsquoShe read this bookrsquo1Polish verbs mark grammatical gender in the past-tense forms In the singular forms the distinction is threefold

masculine (-Oslash) feminine (-a) and neuter (-o) in the plural in contrast the distinction is made between animatepersonal and virile masculine (-i) and non-virile masculine feminine and neuter (-y) We will mark the plurals inthe glosses with the forms M and F respectively but the presented distinction should be kept in mind

Albert Ventayol Boada 130

If we compare (1a and b) we realise that effectively we find the vowel e when there is also a clitic

that marks person yet when it does not attach itself to the verb the epenthesis disappears The

third person forms (1c and d) also confirm this fact the masculine clitics take a phonologically

empty form while marking of the feminine is expressed by the vowel a We will come back to

this phenomenon and other phonological considerations later

What is particular about these verbal clitics is that they can dissociate themselves from

the verb and attach themselves to any phonologically strong element in lineal terms further

to the left than the verb Franks amp King (2000) use (2) from Mikos amp Moravcsik (1986) as a

demonstration of this phenomenon in (2a) the verbal clitic that marks the second person singular

is shifted to the first element from which the sentence is formed while the contrast between (2b

and c) demonstrates the impossibility of clitics to appear further to the right than the verb

(2) (a) Bliski=s mi był jak brat

[closeM=2SG DAT1SG bePRFMSC like brotherNOM]

lsquoYou were as close to me as a brotherrsquo

(b) Kupili=smy lustro

[buyPRFM=1PL mirrorACC]

lsquo(We) bought a mirrorrsquo

(c) Kupili lustro=smy

[BuyPRFM mirrorACC=1PL]

Another example that the authors cite from Mikos amp Moravcsik (1986) is (3) in which the

capacity of Polish verbal clitics to break constituents is made clear while in (3a) the clitic takes

the second element of the chain as its host which is definitely the final element of the first

constituent in (3b) the clitic attaches itself phonologically to the first element thus breaking the

constituent

(3) (a) Czarnego psa=s widział

[blackACC dogACC=2SG seePRFM]

lsquoDid you see a black dogrsquo

(b) Czarnego=s psa widział

[blackACC=2SG dogACC seePRFM]


We have already seen however that this mobility of clitics brings about some consequences at

the phonological level such as the appearance of epenthetic vowels in the case of the masculine

clitics Franks amp King (2000) bring forward two more phonological phenomena to be considered

and that makes them special from the perspective of Spencer and Luiacutes In this way verbal clitics

in Polish count when establishing the end-of-word boundary that determines the phonological

process of vowel raising the phoneme o becomes [u] in the final syllable orthographically

marked oacute If we compare the examples of (4) we see that vowel raising is produced with a verb

in masculine third person (4a) as both the personal marker and the gender morpheme have a

zero realisation and the vowel o ends up blocked in the final syllable This does not happen

in the third person feminine (4b) as the syllable in which we find the vowel o is not final it

is followed by the morpheme of the feminine gender There is no vowel raising either in first

person masculine (4c) because the epenthesis and the clitic have to be taken into account

(4) (a) Moacutegł lsquohe couldrsquo

(b) Mogła lsquoshe couldrsquo

(c) Mogłem lsquoI couldrsquo

This phenomenon of vowel raising makes the verbal clitics of Polish more similar to affixes

rather than to syntactically independent words or elements Despite the previously mentioned

facts the phenomenon of vowel alternation indicates the opposite The following examples

which the authors extract from Rappaport (1988) show how the alternation of a with e which is

produced when the first vowel is blocked in a non-final syllable (5a contrasted with 5b) is not

produced when the final syllable is formed by a clitic and epenthesis (5c)2

(5) (a) Wział lsquohe tookrsquo

(b) Wzieła lsquoshe tookrsquo

(c) Wziałem lsquoI tookrsquo

The second phonological phenomenon that Franks amp King (2000) contribute is the distribution

of stress This phenomenon is relevant insofar as Polish is a fixed-stress language the stressed2The authors note that colloquially the form wziełem is common which would imply an oral tendency towards

the regularisation of forms and rules However they do not bring forward real data on its use


syllable is always the penultimate According to the authors what makes the distribution of

stress peculiar is the fact that the clitics that mark singular persons are kept in mind when

establishing the syllable count (compare 6a with 6b) while their plural counterparts are not (7a

contrasted with 7b)

(6) (a) Kiedy zobaCZYł-e=s dokumenty

[When seePRF-EP=2SG documentsACC]

lsquoWhen did you see the documentsrsquo

(b) Kiedy=s zoBAczył dokumenty

[When=2SG seePRFM documentsACC]

(7) (a) Kiedy zobaCZYły=smy dokumenty

[When seePRFF=1PL documentsACC]

lsquoWhen did we see the documentsrsquo

(b) Kiedy=smy zobaCZYły dokumenty

[When=1PL seePRFF documentsACC]

In this way while the stress changes the syllable in (6b) with regard to (6a) due to the loss of the

clitic this change does not occur in (7b) as in fact the stress in (7a) falls on the antepenultimate

syllable and not the penultimate as would be expected

What these examples show as with (4) and (5) is that Polish verbal clitics are not taken

into consideration in the same way by the hosts that they attach themselves to in some cases one

clitic can enter the syllable count of a word opening up a particular phonological process while

in other cases this same clitic might not be counted and therefore may not trigger phonological

processes that are produced at its level All in all then as put forward by Spencer amp Luiacutes (2012)

Polish verbal clitics have characteristics of both affixes and of words

Other data contributed by Franks amp King (2000) concerning Polish verbal clitics make

reference to omission To this end the authors cite (8) from Mikos amp Moravcsik (1986) as

a prototypical example the person markers appear only once and are recovered from context

in the rest of the cases Yet they do not make remarks as to whether the omission presents

restrictions of any kind The only consideration they make about this point is the possibility that

no personal marker appears (9) from Pruska (1991) provided that the personal pronoun does

not appear in the nominative case this would result in ungrammaticality


(8) Czytali=smy pisali i studiowali

[ReadPRFM=1PL writePRFM and studyPRFM]

lsquoWe read(we) wrote and (we) studiedrsquo

(9) ndash Widzieli=scie ten film

[SeePRFM=2PL thisACC filmACC]

lsquoDid you see this filmrsquo

ndash Tak(my) widzieli ale (my) stwierdzili

ze nic nie wart

[Yes (1PLNOM) watchPRFM but (1PLNOM) concludePRFM

that nothingNOM not valuableNOM]

lsquoYes we saw it but we came to the conclusion that it was not worth watchingrsquo

The example (9) shows precisely that the omission of person markers can be complete in

responses as the context defines the grammatical person Making use of the personal pronouns

in the nominative case however also obliges us to make use of the markers This is quite

surprising if we keep in mind the fact that Polish is a pro-drop language we can do without the

personal pronouns and at the same time the person markers while the use of these pronouns

forces the appearance of these markers Nonetheless the authors do not seem to always be

clear about the spontaneity of this omission (10) from Banski (1997) In this way they relate

this phenomenon to the positioning of the stress the omission of the clitic and the loss of the

epenthesis in (10a) lead to displacement of the stress which does not happen with the feminine

forms (10b) In any case they do not bring forward real data in use which means the question is

yet to be solved

(10) (a) Poszedł=es i zobaCZYł=es zoBAczył

[goPRFM=EP2SG and seePRFM=EP2SG seePRFM]

lsquo(You) went and saw (it)rsquo

(b) Poszła=s i zobaCZYła=s zobaCZYła

[goPRFF=2SG and seePRFF=2SG seePRFF]



Finally the last note that Franks amp King (2000) contribute regarding verbal clitics is the relation-

ship with the copula They give an example for a minimal pair in (11a) we find the adjective

in the form of the masculine plural with the person-marking clitic while in (11b) we find the

copula with the adjective According to the authors we must consider the copula as a whole

basically because of a question of stress the accent falls on the penultimate syllable

(11) (a) ZmeCZEni=scie

[tiredM=2PL]

lsquoYou are tiredrsquo

(b) JeSTEscie zmeczeni

[bePRS2PL tiredM]

Nevertheless when we find the person-marking clitic without the copula the authors note that in

the first and second plural person the suppletive form of the verb byc (corresponding to the third

person of the plural) must be added They put forward the example in (12) from Pruska (1992)

Nevertheless despite indicating that the singular form is impossible they do not offer further

information They just hypothesise that the the use of the copula might have something to do

with the notion of plural In other words the form lsquojestrsquo would be used with singular subjects

whereas the form lsquosarsquo would appear in plural subjects

(12) my ktoacuterzy=smy sajest sługami bozymi

[1PLNOM whichM=1PL bePRS3PLbePRS3SG servantINSPL ofGodINSPL]

lsquo We who are servants of Godrsquo

Thus having provided this typology of phenomena and examples the authors come to the

following conclusions a) that the person-marking verbal clitics are enclitics (b) that the

person-marking singulars are adjunct inflections with the verb and clitics elsewhere (c) that the

person-marking plurals are inflections or clitics wherever they are (d) that the verbal clitics can

break constituents up to the point where the elements of the constituent may not be adjacent

All in all these conclusions suggest that the system of clitics in Polish is found in

a state of flux in other words we can identify in them characteristics found both in affixes

and syntactically independent elements This brings to the fore the idea that Polish clitics are


undergoing a process that will eventually turn them into purely inflective elements This would

actually be the most logical result if we keep in mind that these elements correspond to functional

categories

2 Methodology

With the aim of checking the conclusions drawn by Franks amp King (2000) and of putting to test

our hypothesis (namely that Polish clitics are in a process of grammaticalisation) an experiment

has been designed that gathers a large part of the theoretical framework described in (sect1) The

experiment is made up of two tasks firstly deciding whether a number of provided phrases are

natural or not and secondly deciding which the stressed syllable is in one of the words of a

sentence Both tasks are found mixed throughout the experiment although the first one has a

more significant weight within the experiment as a whole 37 value judgements along with 6

stress judgements There is a reason behind this deciding the stressed syllable in a word only

sheds light on one of the phenomena that are produced when attaching a clitic to a word

In this way the value judgements have allowed for the inclusion of a more extensive

examination of cases of the described phenomena These cases have been divided into six

blocks the six entries in the corpus of annexes The first one (i) holds the phenomenon of

vowel alternation (a gt e) in relation to the displacement of clitics The second and third blocks

(ii-iii) in contrast try to shed light on the acceptability of clitics outside of the verb as well as

the possibility of them breaking constituents The forth and the fifth ones (iv-v) present their

possibility to break constituents and the sixth (vi) deals with the dichotomy between these

elements and the copula

The instructions used when carrying out the tasks were clear and concrete It was not

asked whether something was grammatically correct or not rather it was the naturalness of the

provided phrases that was to be evaluated The options to choose from as value judgements were

ldquoyes this sounds naturalrdquo and ldquono this does not sound naturalrdquo So as to leave the door open to

the considerations of the speakers themselves a third option was also included ldquootherrdquo with an

empty box to fill up This third option seemed relevant to include so as to gather impressions

that the experiment could have suggested at some points It must be kept in mind that in fact a

large portion of the phrases that were included in the task are examples brought forward in (sect1)

by Franks amp King (2000) The aim was to contrast the assertions of the authors bring forward

data of the phenomena they did not offer and shed light on the phrases which do not depend on

the consideration of the speaker This third option allowed thus for the gathering of reflections

on the non-naturalness of some phrases that were not related to clitics


As for the judgements on stress the method of response was always a textbox which was to

be filled with the syllable that each speaker considered to be the stressed one This word was

always either the verb of the phrase or the interrogative particle whether or not the word under

consideration had a singular or plural person-marker attached as an enclitic The motive of this

was to see whether all the speakers applied the same criteria when establishing word boundaries

and whether clitics were part of the syllable count or not

When establishing the sample used in the study the only two variables that were taken

into considerations were (a) speaking Polish as L1 and (b) being originally from Poland

However we tried to cover the largest age range possible in the sample as checking the identity

of the elements in question in different age groups is necessary in order to shed light on the

existence of the on-going process Finally although speakers were asked for their level of

education this variable was not seen as decisive when dealing with the results nor was the

gender of the participants considered to be a relevant variable

The starting number of participants was seventy-four but two of them were excluded for

not complying with the basic premises of the study Of the 72 people interviewed 22 were men

(306) and 50 were women (694) As for the age groups 11 participants belonged to the 16

to 20-year age range (153) 39 belonged to the 21-25 range (528) 14 to that of 26-30 years

(194) 5 participants to that of 31-50 years (69) and 3 were in the range of 51-65 years

(42) With respect to education one of individuals interviewed had gone through primary

schooling (14) 32 had gone through post-obligatory secondary schooling (444) and 38

had gone through education at the tertiary level (528) Regarding the geographical origin

of those interviewed the sample gathers quite a wide panorama Image 1 shows the places of

origin of the participants graphically3

Finally concerning the procedure used in the collection of the data it must be said that

the experiment was designed with the online platform GoogleDrive and was distributed on the

internet using e-mail and social networks (namely Facebook) Despite the limitations they offer

we believe that the results obtained contribute quantitatively significant data

3To be more precise the places of origin of those surveyed were Brzesko Bydgoszcz Dabrowa TarnowskaEłk Gdansk Gdynia Głogoacutew Gołuchoacutew Goacutera Grodzisk Wielkopolski Inowrocław Jawor Katowice KołoKonin Krakoacutew Legnica Lubin Łoacutedz Milcz Olesnica Olsztyn Ostroacutew Wielkopolski Piła Piotrkoacutew TrybunalskiPoznan Sierakoacutew Słupca Słupsk Strzelno Szczecin Srem Swietochłowice Tarnoacutew Torun Tuchoacutew TurekWarszawa Wrocław and Wrzesnia The map was created using Mapbox [23052013]httpsatilesmapboxcomv4hartwallk48e33o9pagehtmlaccess_token=pkeyJ1IjoiaGFydHdhbGwiLCJhIjoiQzdEcnk4NCJ9q57TJvtTvBrGLY-bR4aLMQ75224519748

Data ccopy OpenStreetMap contributors Licensed under the Open Data Commons Open Database License Design ccopyMapbox Licensed according to the Mapbox Terms of Service


Figure 1 Places of origin of those interviewed

3 Results

In this section the results of the experiment designed (sect2) will be presented the discussion of

which being left for the conclusion (sect4) Throughout this section we will follow the distribution

of blocks that the test was based on These blocks correspond to the six entries in the corpus of

annexes (i-vi)

As for the first block which dealt with vowel alternation this alternation is almost never

accepted when the person marker is not found attached to the verb and the vowel ends up blocked

in final syllable (ib) of those surveyed only three out of seventy-two found it natural In contrast

the same context without alternation (ia) gets more support 182 of those surveyed in the

range of 16-30 years 51 of those from 21 to 25 years and 143 of those from 26 to 30 years

This level of acceptance is still low however Nonetheless in both cases one interviewee from

the 31-50 age range considered these sentences to be archaic Finally the third case within the

first block which brings forth the alternation of the verbal vowel and the person-marking enclitic

(ic) was accepted by the speakers in a higher percentage namely 273 of the youngest age

range 154 of those between 21 and 25 years 286 of those between 26 and 30 years and

20 in the range from 31 to 50 years Despite this higher degree of acceptance it is important

to note that three of the participants decided to correct the form offered with one of them saying

that the form found in the test was a terrible error Compare the results of this block in Table 2


Czarnego psa=s wział (ia) Czarnego=s psa wzieł (ib) Czarnego psa wzieł-e=m (ic)

Yes No Other Yes No Other Yes No Other

16-20 182 818 mdash 91 909 mdash 273 727 mdash21-25 51 949 mdash 26 974 mdash 154 795 5126-30 143 857 mdash 71 929 mdash 286 714 mdash31-50 mdash 80 20 mdash 80 20 20 80 mdash51-65 mdash 100 mdash mdash 100 mdash mdash 100 mdash

Table 2 Results for the vowel alternation

Regarding the second and third blocks which deal with the breaking of constituents the mobility

of clitics and the relationship between them the results provide interesting data On the one

hand no speaker accepts the positioning of the person marker more to the right than the verb

(in lineal terms) as natural (iidh iiig) In the same way only one of the seventy-two speakers

accepts prepositions as hosts for clitics (iiib) At the other extreme the phrases that are most

widely considered natural are those in which the person markers take the verb as their host (iic

iiia) in both cases there is between an 80 and 100 level of acceptance in each age range

These parameters are left by the oldest age range in the example of the third block since one of

the speakers rejects the naturalness of the phrase along with the 21-25 age group where the

percentage of acceptance of this particular phrase drops to 641 The reason of such a decrease

may be found in a comment made by one of the participants in this age group according to

which the phrase would sound more natural if the verb was in the first position

With regard to the breaking of constituents the results show that most speakers accept it

especially among the older groups all of the participants between 31 and 65 years of age accept

(iif) as natural while in the three remaining groups the level of acceptance drops to 80 in

the case of (iiie) on the other hand we find 80 of favourable judgements in the older groups

whereas the range of acceptedness drop to 30 in the youngest groups Nonetheless when this

breaking is found in relation with the non-attachment of the person-marker to the verb there are

far fewer judgements in favour both in (iig) and in (iiif) 80 and 100 of those surveyed are

shown to be against it Again the age group that goes from 26 to 30 years old does not show a

strong preference with 50 in favour and 50 against in the case of (iiif) while the results

for (iig) are 357 in favour and 643 against It is important to note the contribution of one

speaker in the second-oldest age group who considers both phrases archaic The results of the

sentences (iifg) and (iiief) can be compared in Table 3


Ciekawa kupiła=s Ciekawa=s kupiła Do mojego nie Do mojego=s nieJanowi ksiazke (iif) Janowi ksiazke (iig) przyszła=s biura (iiie) przyszła biura (iiif)

Yes No Other Yes No Other Yes No Other Yes No Other

16-20 818 182 mdash 364 636 mdash 364 636 mdash 91 909 mdash21-25 795 205 mdash 154 846 mdash 308 692 mdash 179 795 2626-30 786 214 mdash 357 643 mdash 571 429 mdash 50 50 mdash31-50 100 mdash mdash 20 60 20 80 20 mdash 20 60 2051-65 100 mdash mdash mdash 100 mdash 333 667 mdash mdash 100 mdash

Table 3 Results for the breaking of constituents

As far as the examples in (iibe) are concerned the average number of favourable responses

is respective to the total seven In (iiid) the result obtained is identical The only case where

it seems more natural is (iiic) in which the person marker takes the possessive pronoun that

determines the noun of the phrase as a host here the age range from 26 to 30 gives a divided

response with 50 of them in favour of the naturalness of the phrase and 50 against This

proportion is maintained in the next age group (31-50) even though one of the speakers noted

that the phrase sounds archaic

Regarding the phenomena of omission of the fourth and fifth blocks (iv-v) the results

are very clear Whenever there are two verbs if the clitic appears it has to be on both of them

Thus in block (iv) between the 90 and 100 of those surveyed along all the age groups reject

those cases in which the verbal clitic appeared only with one of the two verbs either the first

one or the second one regardless of the gender form of the verb (ivbcgh) This percentage

appears again in those contexts where in addition to a verbal personal marker adjunct to one of

the verbs appears as well as the personal pronoun in nominative (ivde) Compare the results

obtained for the masculine inflected forms in Table 4

Czy poszedłe=s i zobaczyłe=s Czy poszedłe=s i zobaczył Czy poszedłi zobaczyłe=sco sie tam stało (ivf) co sie tam stało (ivg) co sie tam stało (ivh)


16-20 818 182 mdash 91 909 mdash 91 909 mdash21-25 923 77 mdash mdash 100 mdash 26 974 mdash26-30 929 71 mdash 143 857 mdash mdash 100 mdash31-50 100 mdash mdash mdash 100 mdash mdash 100 mdash51-65 100 mdash mdash mdash 100 mdash mdash 100 mdash

Table 4 Results for the phenomenon of omission in masculine inflected verbal forms

The block (v) on the other hand provides very similar data A number between 90 and 100

of those surveyed in every age group reject the appearance of a single person marker whether or

not this single mark is that of the medial verb (ve) or that of the final verb (vf) The percentage


of acceptance when this single mark appears adjoined to the first verb (vd) however is higher

125 of the total (the highest percentage of acceptance is found in the oldest group where 33

of those surveyed come out in favour of it) Regarding the context in which the strong form of

the personal pronoun appears but the person marker does not (vb) between 80 and 100 of

the speakers do not consider it natural Nonetheless it is interesting to note that within the age

group that covers 26 to 30 years 375 consider the phrase natural and sounds good which is

quite surprising given that this context was believed to be ungrammatical

Finally it is important to note that the lack of a person marker and strong personal

pronoun in a response in a predetermined context (vc) is accepted by 153 of the participants

with the highest level of support being among those in the age range of 26-30 years at 357 On

the other hand 389 of those surveyed perceive phrases with a single enclitic person marker in

the pronoun in nominative in initial position (va) to be natural Yet the differences here become

notable when the groups of the study are compared as there is a greater level of rejection of the

phrase observed when the age of the participants is lower it is considered natural by 667

60 and 572 of the three oldest groups (from oldest to youngest) while the acceptance of

the phrase among those between 21 and 25 years of age is 282 and among those between 16

and 21 years of age 364 However at least two of the comments made by participants in the

survey should be added although they accepted the naturalness of the response they think that

the question was not well-formed and that the imperfective aspect would be more relevant

With regard to the block (vi) which deals with the dichotomy between person markers

and the copula the results are quite categorical Thus all of the speakers reject cases where the

plural person marker appears attached to an adjective and the form of the copula that we find is

the singular (vibd) Despite this the level of naturalness that the speakers find in the contexts

in which it is the suppletive plural form of the copula that acts as an auxiliary is very low More

precisely for (vic) only 153 of the total come out in favour of it while for (via) the percentage

of positive judgements is 125 It is relevant to note that it is precisely the youngest age group

that considers these two phrases to be natural 364 and 182 respectively Compare the

results in Table 5


Zmeczeni=scie sa Zmeczeni=scie jest ndash Jak sie macie ndash Jak sie macie(via) (vib) ndash Zmeczeni=scie ndash Zmeczeni=scie jest

sa (vic) (vid)


16-20 182 636 182 mdash 100 mdash 364 636 mdash mdash 100 mdash21-25 77 923 mdash mdash 100 mdash 103 871 mdash mdash 100 mdash26-30 214 786 mdash mdash 100 mdash 214 786 mdash mdash 100 mdash31-50 mdash 80 20 mdash 100 mdash mdash 80 20 mdash 100 mdash51-65 333 667 mdash mdash 100 mdash mdash 100 mdash mdash 100 mdash

Table 5 Results for the dichotomy between peson markers and the copula

Before finishing this section the results obtained in the stress judgements should also be

discussed On the one hand as far as cases in which the stressed syllable of the interrogative

pronoun (i) are concerned the results show that the speakers maintain the stress in the original

syllable even when the interrogative pronoun has a plural person marker as an enclitic However

differences were observed between the different age groups the range between 51 and 65 years

of age always maintains the stressed syllable while among the younger groups such as the

range from 21 to 25 years only 565 of those surveyed On the other hand when the stressed

syllable of the verb was asked for (ii) the non-attachment of the singular person-marking to the

verb (iib) represents in all the age groups a displacement of the stress to the directly anterior

syllable with respect to the cases where the singular person marker is found attached to the verb

(iia)

In relation to the plural person markers however relevant differences according to the

study groups are observed 667 of the oldest age range and 40 of the one just below put

the stress on the same syllable whether (iic) or not (iid) the person marker takes the verb as a

host which indicates that the clitic is not counted when the penultimate syllable of the verb is

determined in order to decide where the stress must fall However the results obtained for the

age group from 26 to 30 years show that 716 of those surveyed always stress the penultimate

syllable so a displacement of the stress in cases where the clitics take the verb as a host is found

Among those youngest this difference is not so proeminent in the sixteen-to-twenty-year-old

group only 454 count the clitic when establishing the word boundary while among those

between 21 and 26 years of age the percentage rises to 461


4 Conclusion

After analysing the results from the experiment we can assert that our hypothesis remains veri-

fied at least in part Thus the data gathered show as a matter of fact that Polish native speakers

generally accept the adjunction of clitic personal markers to verbs whereas the naturalness of

the adjunction of these elements to other hosts decreases as the age of the participants does

Therefore it is patently clear that these elements are becoming rather inflectional morphemes

The lack of more participants in the eldest age group or even the inclusion of more target groups

in the age range makes it difficult to tell whether this process may have not already finished

Actually the fact that some of the phenomena described by Franks amp King (2000) were not

accepted as natural suggests that the loss of properties typical of syntactically independent words

is gradual Thus the data confirmed that the clitic personal markers cannot be attached to a

word that is further to the right than the verb in linear terms In addition to that the impossibility

for such clitics to take prepositions as hosts was also confirmed On the other hand the cases

regarded as most acceptable and natural are those where the clitic personal marker takes personal

pronouns in nominative case as a host

Regarding the phenomenon of omission the results are clear none of the groups

considered sentences with more than one verb and a single personal marker natural (contra

Franks amp King (2000)) Accordingly the contexts in which no clitic personal marker appeared

throughout the whole proposition were not regarded as natural either even if such a proposition

was a concise answer to a question that had established a contextual and pragmatic framework

These data again contradict the description of the authors cited From our results it can only be

deduced that omission is only accepted if at least one personal marker appears adjunct to the

personal pronoun in nominative case in sentence initial position With respect to clitic personal

markers and their relationship with the copula on the other hand it can be asserted that the

co-occurrence of the singular copula with plural personal markers is not acceptable Surprising

as it may seem it is the younger group that argues more for the naturalness of those cases in

which the suppletive form of the copula appears since such sentences are no longer regarded as

natural among the elder groups

Finally the results of the stress judgements display a tendency among the youngest to

always stress the penultimate syllable of the verb thus the verbal personal markers are no longer

seen as clitics Nevertheless this is not true for those cases where the clitic is adjunct to a

non-verbal host Here the clitics are perceived as such and are not included when establishing

the syllable count


All in all in order to be critical with our own work the door should be left open to the pragmatic

context not considered in our study Actually considering the contexts presented here as natural

does not imply necessarily that they are pragmatically neutral That notwithstanding none of

the typological descriptions we have referred to has included such a possibility as a possible

explanation Thus we are positive this variable should be considered in further studies along

with the considerations made here ie the existence of an ongoing process of grammaticalisation

of the personal verbal markers

References

Banski Piotr (1997) lsquoPolish auxiliary clitics morphology or syntaxrsquo In ZAS Papers in

Linguistics vol 6 17ndash27 Berlin Zentrum fuumlr Allgemeine Sprachwis-

senschaft

Franks Steven amp King Tracy H (2000) A Handbook of Slavic Clitics New York Oxford

University Press

Mikos Michael J amp Moravcsik Edith A (1986) lsquoMoving clitics in Polish and some crosslin-

guistic generalizationsrsquo In Studia Slavica 32 327ndash336

Pruska Bozena (1991) Polish mobile inflection San Diego University of California

Pruska Bozena (1992) lsquoMobile morphemes and agreement in Polishrsquo In Working Papers in

Linguistics - University of Washington 10 107ndash121

Rappaport Gilbert (1988) lsquoOn the relationship between prosodic and syntactic properties of the

pronouns in the Slavonic languagesrsquo In A Schenker ed American

Contributions to the Tenth International Congress of Slavists 301ndash327

Columbus Slavica

Spencer Andrew amp Luiacutes Ana R (2012) Clitics An Introduction New York Cambridge

University Press

Swan Oscar E (2002) A Grammar of Contemporary Polish Bloomington Slavica Publishers


Annexes

The following pages collect the sentences used in the judgement value test the results of which

were presented in (sect3) The sentences are grouped in blocks according to the phenomenon

exhibited A single translation is provided for each block since the sentences only differ in

terms of clitic adjunction

(i) (a) Czarnego psa=s wział

[blackACC dogACC=2SG takePRFM]

lsquoDid you (M) take the black dogrsquo

(b) Czarnego=s psa wzieł

[blackACC=2SG dogACC takePRFM]

(c) Czarnego psa wzieł-e=m

[blackACC dogACC takePRFM-EP=1SG]

(ii) (a) Ciekawa=s ksiazke kupiła Janowi

[InterestingACC=2SG bookACC buyPRFF JohnDAT]

lsquo(YouF) bought an interesting book to Johnrsquo

(b) Ciekawa ksiazke=s kupiła Janowi

[InterestingACC bookACC=2SG buyPRFF JohnDAT]

(c) Ciekawa ksiazke kupiła=s Janowi

[InterestingACC bookACC buyPRFF=2SG JohnDAT]

(d) Ciekawa ksiazke kupiła Janowi=s

[InterestingACC bookACC buyPRFF JohnDAT=2SG]

(e) Ciekawa ksiazke Janowi=s kupiła

[InterestingACC bookACC JohnDAT=2SG buyPRFF]

(f) Ciekawa kupiła=s Janowi ksiazke

[InterestingACC buyPRFF=2SG JohnDAT bookACC]

(g) Ciekawa=s kupila Janowi ksiazke

[InterestingACC=2SG buyPRFF JohnDAT bookACC]

(h) Ciekawa kupiła Janowi ksiazke=s

[InterestingACC buyPRFF JohnDAT bookACC=2SG]


(iii) (a) Do mojego biura nie przyszła=s

[PRP POSS1SGGEN offficeGEN not comePRFF=2SG]

lsquo(YouF) did not come to my officersquo

(b) Do=s mojego biura nie przyszła

[PRP=2SG POSS1SGGEN officeGEN not comePRFF]

(c) Do mojego=s biura nie przyszła

[PRP POSS1SGGEN=2SG officeGEN not comePRFF]

(d) Do mojego biura=s nie przyszła

[PRP POSS1SGGEN officeGEN=2SG not comePRFF]

(e) Do mojego nie przyszła=s biura

[PRP POSS1SGGEN not comePRFF=2SG officeGEN]

(f) Do mojego=s nie przyszła biura

[PRP POSS1SGGEN=2SG not comePRFF officeGEN]

(g) Do mojego nie przyszła biura=s

[PRP POSS1SGGEN not comePRFF officeGEN=2SG]

(iv) (a) Czy poszła=s i zobaczyła=s co tam sie stało

[Q goPRFF=2SG and seePRFF=2SG whatACC there REFL happenPRFN]

lsquoDid you (F) go there and see what had happenedrsquo

(b) Czy poszła=s i zobaczyła co tam sie stało

[Q goPRFF=2SG and seePRFF whatACC there REFL happenPRFN]

(c) Czy poszła i zobaczyła=s co tam sie stało

[Q goPRFF and seePRFF=2SG whatACC there REFL happenPRFN]

(d) Czy Ty poszła=s i zobaczyła co tam sie stało

[Q 2SGNOM goPRFF=2SG and seePRFF what=ACC there REFL happenPRFN]

(e) Czy Ty poszła i zobaczyła=s co tam sie stało

[Q 2SGNOM goPRFF and seePRFF=2SG what=ACC there REFL happenPRFN]

(f) Czy poszedłe=s i zobaczyłe=s co tam sie stało

[Q goPRFMEP=2SG and seePRFMEP=2SG what=ACC there REFL hap-

penPRFN]

lsquoDid you (M) go there and see what had happenedrsquo

(g) Czy poszedłe=s i zobaczył co tam sie stało

[Q goPRFMEP=2SG and seePRFM what=ACC there REFL happenPRFN]


(h) Czy poszedł i zobaczyłe=s co tam sie stało

[Q goPRFM and seePRFMEP=2SG what=ACC there REFL happenPRFN]

(v) (a) - Co zrobili=scie wczoraj w bibliotece

[what=ACC doPRFM=2PL yesterday PRP libraryLOC]

- My=smy czytali pisali i studiowali

[1PLNOM=1PL readPRFM writePRFM and sturyPRFM]

lsquo- What did you do yesterday in the library - We read wrote and studiedrsquo

(b) - Co zrobili=scie wczoraj w bibliotece


- My czytali pisali i studiowali

[1PLNOM readPRFM writePRFM and studyPRFM]

(c) - Co zrobili=scie wczoraj w bibliotece


- Czytali pisali i studiowali

[readPRFM writePRFM and studyPRFM]

(d) - Co zrobili=scie wczoraj w bibliotece


- Czytali=smy pisali i studiowali

[readPRFM=1PL writePRFM and studyPRFM]

(e) - Co zrobili=scie wczoraj w bibliotece


- Czytali pisali=smy i studiowali


(f) - Co zrobili=scie wczoraj w bibliotece


- Czytali pisali i studiowali=smy

[readPRFM writePRFM and studyPRFM=1PL]

(vi) (a) Zmeczeni=scie sa

[TiredM=2PL AUXbePRS3PL]

lsquoAre (YouM) tiredrsquo


(b) Zmeczeni=scie jest

[TiredM=2PL AUXbePRS3SG]

(c) - Jak sie macie

[How REFL havePRS2PL]

- Zmeczeni=smy sa

[TiredM=1PL AUXserPRS3PL]

lsquo- How are you - (WeM) are tiredrsquo

(d) - Jak sie macie


- Zmeczeni=smy jest

[TiredM=1PL AUXserPRS3SG]

Finally the ensuing sentences are those used in the stress judgement test Regarding

the first two sentences the speaker was asked to consider which the stressed syllable of the

interrogative pronoun was On the other hand the remaining sentences were used to ask inquire

about the stressed syllable of the verb

(i) (a) Czego nie zrobili=scie

[whatGEN not doPRFM=2PL]

lsquoWhat did not (M) you dorsquo

(b) Czego=scie nie zrobili

[whatGEN=2PL not doPRFM]

(ii) (a) Gdzie zobaczyłe=s te dokumenty

[where seePRFMEP=2SG theseACC documentsACC]

lsquoWhere did you (M) see these documentsrsquo

(b) Gdzie=s zobaczył te dokumenty

[where=2SG seePRFMEP theseACC documentsACC]

(c) Gdzie zobaczyły=scie te dokumenty

[where seePRFF=2PL theseACC documentsACC]

(d) Gdzie=scie zobaczyły te dokumenty

[where=2PL seePRFF theseACC documentsACC]

Introduction the Polish verbal clitics

Methodology

Results

Conclusion


Preterite (clitic form) Copula (full form)

1st p sg -(e)m jestem2nd p sg -(e)s jestes3rd p sg mdash jest1st p pl -(e)smy jestesmy2nd p pl -(e)scie jestescie3rd p pl mdash sa

Table 1 Verbal clitics in Polish

we consider discursive clitics following the terminology of Spencer amp Luiacutes (2012) However

here we will only revise the first group Table 1 adapted from the same authors summarises the

existing verbal clitics

There are several points in the above table that must be kept in mind First of all the

dichotomy between preterite and copula lies in two facts On the one hand the clitic forms

appear only in correlation with the past tenses of verbs on the other hand the similarity between

verbal clitics and the conjugated forms of the verb byc lsquoto bersquo cannot be mere coincidence In a

general sense it should be kept in mind that the clitics shown in the table act as person markers

ie they introduce information about number and person except for the third persons which

do not have specific forms Finally regarding the e in parentheses it is normally identified as a

marker of the masculine gender1 This however is inaccurate because the nature of this vowel

is epenthetic (Swan 2002)

(1) (a) Ja czytał-e=m te ksiazke

[I readPRF -M=1SG thisACC bookACC]


(b) Ja=m czytał-Oslash te ksiazke

[I=1SG readPRF-M thisACC bookACC]


(c) On czytał-Oslash te ksiazke

[He readPRF-M thisACC bookACC]

(d) Ona czytał-a te ksiazke

[She readPRF-F thisACC bookACC]

lsquoShe read this bookrsquo1Polish verbs mark grammatical gender in the past-tense forms In the singular forms the distinction is threefold

masculine (-Oslash) feminine (-a) and neuter (-o) in the plural in contrast the distinction is made between animatepersonal and virile masculine (-i) and non-virile masculine feminine and neuter (-y) We will mark the plurals inthe glosses with the forms M and F respectively but the presented distinction should be kept in mind

























constituent






































contrasted with 7b)



































ze nic nie wart














yet to be solved














[tiredM=2PL]



[bePRS2PL tiredM]






















categories

2 Methodology






























































3 Results




annexes (i-vi)


































































































results in Table 5



sa (vic) (vid)














(iia)












4 Conclusion
































the syllable count









References



senschaft


University Press









Columbus Slavica


University Press



Annexes


























































penPRFN]






































(c) - Jak sie macie


- Zmeczeni=smy sa



(d) - Jak sie macie


- Zmeczeni=smy jest





















Methodology

Results

Conclusion

























constituent






































contrasted with 7b)



































ze nic nie wart














yet to be solved














[tiredM=2PL]



[bePRS2PL tiredM]






















categories

2 Methodology






























































3 Results




annexes (i-vi)


































































































results in Table 5



sa (vic) (vid)














(iia)












4 Conclusion
































the syllable count









References



senschaft


University Press









Columbus Slavica


University Press



Annexes


























































penPRFN]






































(c) - Jak sie macie


- Zmeczeni=smy sa



(d) - Jak sie macie


- Zmeczeni=smy jest





















Methodology

Results

Conclusion

































contrasted with 7b)



































ze nic nie wart














yet to be solved














[tiredM=2PL]



[bePRS2PL tiredM]






















categories

2 Methodology






























































3 Results




annexes (i-vi)


































































































results in Table 5



sa (vic) (vid)














(iia)












4 Conclusion
































the syllable count









References



senschaft


University Press









Columbus Slavica


University Press



Annexes


























































penPRFN]






































(c) - Jak sie macie


- Zmeczeni=smy sa



(d) - Jak sie macie


- Zmeczeni=smy jest





















Methodology

Results

Conclusion









ze nic nie wart














yet to be solved














[tiredM=2PL]



[bePRS2PL tiredM]






















categories

2 Methodology






























































3 Results




annexes (i-vi)


































































































results in Table 5



sa (vic) (vid)














(iia)












4 Conclusion
































the syllable count









References



senschaft


University Press









Columbus Slavica


University Press



Annexes


























































penPRFN]






































(c) - Jak sie macie


- Zmeczeni=smy sa



(d) - Jak sie macie


- Zmeczeni=smy jest





















Methodology

Results

Conclusion








[tiredM=2PL]



[bePRS2PL tiredM]






















categories

2 Methodology






























































3 Results




annexes (i-vi)


































































































results in Table 5



sa (vic) (vid)














(iia)












4 Conclusion
































the syllable count









References



senschaft


University Press









Columbus Slavica


University Press



Annexes


























































penPRFN]






































(c) - Jak sie macie


- Zmeczeni=smy sa



(d) - Jak sie macie


- Zmeczeni=smy jest





















Methodology

Results

Conclusion




categories

2 Methodology






























































3 Results




annexes (i-vi)


































































































results in Table 5



sa (vic) (vid)














(iia)












4 Conclusion
































the syllable count









References



senschaft


University Press









Columbus Slavica


University Press



Annexes


























































penPRFN]






































(c) - Jak sie macie


- Zmeczeni=smy sa



(d) - Jak sie macie


- Zmeczeni=smy jest





















Methodology

Results

Conclusion

































3 Results




annexes (i-vi)


































































































results in Table 5



sa (vic) (vid)














(iia)












4 Conclusion
































the syllable count









References



senschaft


University Press









Columbus Slavica


University Press



Annexes


























































penPRFN]






































(c) - Jak sie macie


- Zmeczeni=smy sa



(d) - Jak sie macie


- Zmeczeni=smy jest





















Methodology

Results

Conclusion





















































































results in Table 5



sa (vic) (vid)














(iia)












4 Conclusion
































the syllable count









References



senschaft


University Press









Columbus Slavica


University Press



Annexes


























































penPRFN]






































(c) - Jak sie macie


- Zmeczeni=smy sa



(d) - Jak sie macie


- Zmeczeni=smy jest





















Methodology

Results

Conclusion



sa (vic) (vid)














(iia)












4 Conclusion
































the syllable count









References



senschaft


University Press









Columbus Slavica


University Press



Annexes


























































penPRFN]






































(c) - Jak sie macie


- Zmeczeni=smy sa



(d) - Jak sie macie


- Zmeczeni=smy jest





















Methodology

Results

Conclusion


4 Conclusion
































the syllable count









References



senschaft


University Press









Columbus Slavica


University Press



Annexes


























































penPRFN]






































(c) - Jak sie macie


- Zmeczeni=smy sa



(d) - Jak sie macie


- Zmeczeni=smy jest





















Methodology

Results

Conclusion









References



senschaft


University Press









Columbus Slavica


University Press



Annexes


























































penPRFN]






































(c) - Jak sie macie


- Zmeczeni=smy sa



(d) - Jak sie macie


- Zmeczeni=smy jest





















Methodology

Results

Conclusion


Annexes


























































penPRFN]






































(c) - Jak sie macie


- Zmeczeni=smy sa



(d) - Jak sie macie


- Zmeczeni=smy jest





















Methodology

Results

Conclusion






























penPRFN]






































(c) - Jak sie macie


- Zmeczeni=smy sa



(d) - Jak sie macie


- Zmeczeni=smy jest





















Methodology

Results

Conclusion



































(c) - Jak sie macie


- Zmeczeni=smy sa



(d) - Jak sie macie


- Zmeczeni=smy jest





















Methodology

Results

Conclusion

What do we learn from clitics? The case study of Polish

Documents

Transcript of What do we learn from clitics? The case study of Polish